Item

Abstract

Under many conditions, the human visuomotor system quickly adapts when confronted with spatially discrepant stimuli. This adaptability can easily be demonstrated with for example
the change in pointing behavior when rst wearing prism glasses. Here we asked whether
the quality of feedback, that is its reliability, has an effect on the rate of adaptation. The hypothesis
was that the system should adapt more quickly if the feedback was more reliable. To
investigate this question we conducted two adaptation experiments: One pointing experiment
(closed-loop), in which we measured the rate of adaptation to a lateral prismatic displacement,
and a grasping experiment (open-loop), in which we measured adaptation to a size conict of
visually magnied objects. The experiments were conducted in three phases: A pre-adaptation
phase to establish baseline performance, an adaptation phase in which the visuomotor conict
was introduced, and a post-adaptation phase to determine the after-effect. We determined the
rate of adaptation from the change in visuomotor behavior during the second and third phases.
In several conditions the reliability of feedback was manipulated. In the pointing experiment
we varied the feedback reliability by blurring the target stimulus, which was a Gaussian blob
of 10 contrast (sigma = 4, 32, 48 deg). In the grasping experiment we presented rectangular
blocks of ve different sizes for open-loop grasping. Here the feedback was manipulated by
either adding noise to the visual display or by varying the haptic presentation time. The effectiveness
of all feedback manipulations was assessed by measuring discrimination thresholds.
To our surprise, in both experiments, closed-loop pointing and open-loop grasping, there was
no signicant effect of the reliability of feedback on the rate of adaptation. We conclude that
the rate of visuomotor adaptation depends on the sensory estimate that determines the conict
but not on its reliability.